Look at @OstiumLabs, I rarely focus on TVL first.

It's not that TVL is useless, but its explanatory power here is really limited.

What truly makes me stop and think is the trading volume structure, especially the source of open interest (OI) — in a derivatives protocol, 95% of OI comes from traditional assets, which itself is very telling.

This structure can't be faked, nor can it naturally form during airdrop periods; fundamentally, it's about filtering users:
Some use Ostium not for on-chain sentiment betting, but for real-world judgments outside crypto—like interest rates, indices, macro trends.

This is very rare in RWA projects — most are "narrative first," using concepts to cover up real usage;
Ostium, on the other hand, starts with tangible use cases and gradually reveals the narrative.

There's also a common misinterpretation:
Ostium's TVL isn't high, not because the funds are small, but because the money simply doesn't need to enter the platform.
During trading, funds stay entirely in users' wallets, with no "deposit — lock-up — retention" step.

Therefore, TVL is more like a deliberately suppressed indicator; low doesn't mean weak, but rather shows it doesn't rely on fund retention to create a sense of security.

Looking at the real OI structure, the proportion of traditional assets, and the non-custodial fund flow together, you realize Ostium has never aimed to appear "big," but rather to be a long-term reliable judgment tool for users.

This is also one of the few projects I seriously follow in the RWA track.
The S2 score has started; those who haven't participated can do some research.